New Mexico Tech UAV Sr. Design Project

[Tom Harper]

Last year I proposed a low cost UAV controller with Camera control as a Sr, Project. The following students signed up for the project (see photo):


Jamison London
Chris Whited
Mike Zaw
Mike Pursley
Steven Myers
Karina Munoz


Dr. Hector Erives manages the Sr. Projects and Dr. Rene Arechiga is the team advisor. I supply the airplane(s) and Carl is the Pilot.

I met with the team today and was very pleased with the progress. Their finished PC board is shown in the photo. Looks like the interface is compatible with my wiring. The blue wires are a patch for a 'to be delivered' voltage regulator.

This solution uses two MSP430 microcontrollers. One for the auto/manual control and a second for the navigation algorithm. Waypoints are programmed in the model using the 8 line JTAG connector. The board is 2 1/2 inches square. There is no external crystal or oscillator. The cost of the two processors is about $7. The code is in 'C' and is well documented.

The software is not quite complete. We have scheduled the first flight for 29 Mar.

Great way to get the best and brightest working on our project.

[tnavressdog]

So how did the test flight go?

[Tom Harper]

The team says they are having trouble with the UART, so they postponed the test flight. Real time processing in 'C' is not easy.

They should give me a new date this week. I'll post their progress.

Thanks for the interest.

Tom

[tnavressdog]

Great. I'm looking forward to the report. What aircraft are you guys using and any pictures of the complete setup? Sounds great.

[Tom Harper]

Latest status from the team:

"Here is an update on our progress. All of the software is integrated and
is in a debugging stage. Currently the main problems are PWM jitter and
serial data reception problems. The jitter seems to be related to the
serial code. Serial data is being received but not an entire message. As
soon as we have these items under control we contact you."

[Tom Harper]

Dropped by the lab yesterday to see the progress. There is a new PC board on the way. In the mean time the team is working with a breadboard. Took it out for a field test on the soccer field. The servo moved some but there were power supply problems.

Looks like they have created some excellent tools. Shouldn't be long now.

Other projects are underway. A satelite antenna and a lens controller were on the bench. The soft drink machine is comming together. Lot's of activity.

[Tom Harper]

Latest message form team leader:


"We managed to work out most of the kinks in the PWM code tonight which
was the last major hurdle. The system has been tested with the rudder as
the only guide between waypoints with complete success. We actually
tried literately running the course which worked much better since the
gps points were farther apart. Finally we ran a new course of points on
the road by driving around campus which also was a success."

[Tom Harper]

The final presentations were last Friday.

The team solved most of the problems and demonstrated a system that performed well in tests. The primary difficulties were associated with the use of 'C' to program a small, real time system. Lot's of problems with PWM stability and code size. However, all of the system requirements were met. A neat PC board with surface mount components and a total system cost (in units) of less than $100. Some very impressive laptop support software.

Haven't flown it yet. But, I think we will. Working on a date.

Tom

[HELModels]

What is it about C that would cause PWM instability? Or what is it about PWM that makes it difficult to control using C? I am no programmer, but I am curious.